4. Regression Modelling
4.3 Model Two
This section compares producers that enter into a high or low level of buy-in to a commitment to one company. Table 23 and Table 24 illustrate in more detail some of the differences between levels of commitment entered into by different producers. Pricing specifications refer to the method by which producers elect to be paid for their lambs. The schedule price is the open market price per kilogram carcase weight that is released weekly by companies. This is usually emailed to farmers that are suppliers of meat companies, or posted on company websites. Producers can be paid premiums above the schedule price based on commitment or meeting other specifications. If producers commit supply or enter contracts, there is the option of fixing the price per kilogram or per head for a certain period of time, or for the year. This option is not generally popular among farmers, as in the past the schedule price has been as high as or higher than the contract price. Some companies offer a minimum price, which can increase if the schedule price increases.
The majority of producers who always commit sell some or all lambs on the schedule price. Seventy- six producers (out of 324 total Committers analysed in this model, or 23%) sell some or all lambs under a fixed price. Another ten producers sell some or all lambs under a minimum price system. Table 23: Commitment pricing specifications
Note: Percentages do not add to 100% as some producer’s ticked more than one option.
When deciding whether to commit or not, producers will need to consider the level of commitment they give to processors. Stringency in enforcing commitment varies between companies, ranging from no action, to losing preferential processing space access, or losing premiums. If a producer breaks commitment, there are generally few specific penalties, as commitments are loosely based around mutual agreement rather than contractual obligations. When committing lamb supply, producers are asked at the start of the season (around October) to provide an estimate of the number of lambs they will be able to commit over the following season. Sometimes an estimate of when these lambs will be ready for slaughter is also required. This may be a monthly or weekly specification. For more stringent programmes, producers may be required to meet specific weight, fat grade, or breeding and feeding specifications. Producers then will attempt to meet these estimated lamb numbers, dates and other requirements. There may be a review of estimated delivery half-way through the season, or at more regular interviews based on a producer’s updated information (e.g. lambing percentages, pasture and lamb growth rates). Companies may provide flexibility in terms of numbers and dates specified. For example farmers may be allowed to supply a week earlier or later than specified, or 10 per cent more or fewer animals than specified in a week without any repercussions. This flexibility is especially critical in years where climatic conditions are extreme or unpredicted. While in some cases the more stringent the requirements, the higher the premiums and rewards on offer, this is not always the case. Some farmers commit lambs to varying degrees of specificity, and receive the schedule price.
Pricing specification Number Proportion of people that commit
Schedule plus premiums 239 74%
Schedule 64 20%
Fixed price per kilogram 61 19%
Fixed price per head 15 5%
Table 24: Commitment delivery specifications
Note: Percentages do not add to 100% as some producer’s ticked more than one option.
The most flexible commitment option is to solely commit an annual number of lambs to one company. While 109 farmers ticked this box, it can be assumed from the previous question that the 324 that ticked always or almost always commit, make an annual commitment. It could however be the case, that some of those 324 only commit lambs for a shorter period of time such as over summer, over winter, or monthly in advance. From those 324 producers, 89 (27%) then specify numbers by month, and 57 (17%) also commit to a weekly number. A minimum delivery of numbers is also a popular commitment option although this would usually be tied in to an annual commitment. Meeting weight range specifications is the next most common commitment option with 80 (24%) of committed producers meeting weight specifications. Fat grades specifications is slightly less prescribed with 52 (16%) farmer aiming to meet this requirement. Feeding and breeding requirements are often linked to very specific supermarket programmes, and a low number of producers opt in to these high requirement options.
The second regression partitioning model Error! Reference source not found. compares producers that commit to a high level (High Committers) with those that have a low level of commitment (Low Committers). A producer was defined as a High Committer (HC) if they ticked that their commitment included higher specifications of price or delivery. These include any of a fixed price (per kilogram or per head), meeting monthly or weekly delivery of lambs, meeting weight and fat specifications, and breeding or feeding specifications. Low Committers (LC) commit to an annual supply arrangement with schedule prices, although they could also be meeting weight specifications, and a minimum number supplied. High committers may also receive schedule prices, but it is one of the higher specifications which determine whether they are High Committers or not. From the 324 Committers, 193 are allocated to the Low Committers, and 131 to the High Committers. This means that at the top branch of the model a producer has a 60 per cent chance of being a Low Committer, and 40 per cent chance of being a High Committer.
The largest cause of differentiation as to whether or not producers are High Committers is based on the producers score for the Market component (Error! Reference source not found.). The Market component is a measure of to what extent the producer is interested in where their lamb is consumed and who they sell to, and changes made on-farm to better meet market requirements. The Market component score has a positive significant correlation with the Influencer, Active and Conscientiousness components scores. This means that the other components could also cause a large degree of differentiation between producers but that their effect is overpowered by the Market component.
Requirements Number Proportion of people that commit
Annual delivery 109 34%
Monthly delivery 89 27%
Minimum number delivered 85 26%
Weight specification 80 25%
Weekly delivery 57 18%
Fat grade specification 52 16%
Feeding requirement 9 3%
Market< 3.407 Conven>=-1.04 Market< -0.777 Auton>=-0.522 Influencer< -0.8861 Age=4,5,6 Conven< 0.7489 Auton>=0.5374 Trusting>=0.9333 0.13 n=78 0.042 n=24 0.27 n=44 n=27 0.52 0.45 n=20 0.75 n=20 0.35 n=26 0.48 n=21 0.74 n=27 0.84 n=37
A
B
C
D
E
F
G
H
I
J
High market Med market Low Convenience Low market Med-High Convenience High Conv Med Conv 0.29 n=213 0.35 n = 287 0.53 n = 74 0.28 n = 95 0.60 n = 40 0.62 n = 48 0.38 n = 135 Low Market 0.40 n = 324 Low Autonomy High AutonomyHigh Autonomy Low Autonomy
Producers are drafted initially by whether or not they have a Market score below 3.407. Producers with a very high Market score enter final Node A, whereby this small number of producers has an 84 per cent chance of being a High Committer. Producers with a lower Market score are then partitioned based on their Convenience score. Producers with high Convenience scores prefer to let an agent take care of selling arrangements, do not see value in comparing different companies or prices, and do not like to spend a lot of time weighing lambs or targeting premiums and weight ranges. If producers have a high Convenience score, they are drafted to the left, where they are then partitioned on their Market score. Those with a very low Market score are allocated to final node B and have only a 13 per cent chance of being a High Committer.
Producers with a low first branch Market score, and a high Convenience score, and then a medium third branch Market score (low in the first branch, but then high in the second Market branch), are then partitioned by their Autonomy score. The Autonomy score is orthogonal to the Convenience score, with the main difference being related to whether or not a producer prefers to use a third party agent or not. Producers with a high Autonomy score at this branch are then partitioned based on their Influencer score. The Influencer score relates to the producers perceived level of bargaining power, status amongst other producers, and willingness to adopt new supply options. If the producer has a low Influencer score at this branch, they are allocated to node C and have almost no chance of being a High Committer. If they have a high Influencer score, they are drafted to the right, and are then partitioned based on their age. Producers in the groups above 50 years of age are partitioned to node D and have a 27 per cent chance of being a High Committer. Those producers under 50 years of age are partitioned to node E and have a 52 per cent chance of being a High Committer.
For producers that have a low original Market score, high Convenience score at the second branch, medium Market and low Autonomy scores, are then partitioned based on their Convenience score again. Those with a medium Convenience score (high in the first Convenience branch, but not high in second Convenience branch) are allocated to Node F and have a 45 per cent chance of being a High Committer. These producers have a higher score for Convenience than those that were partitioned to the right at the first Convenience branch, but are still average to below average (all scores for PCA’s have been standardised and so the average score for each PCA is 0). Those with above average scores for Convenience are allocated to node G and have a 75 per cent chance of being a High Committer. It is of interest to note that producers with high Convenience scores are more likely to be High Committers, but it could be due to the low Autonomy and medium Market scores in combination with a preference for convenience of the buyer deciding that the producer ends up making high commitments. Producers with low Autonomy and medium Convenience scores may have medium scores for the role of the agent and are less worried with autonomy or convenience. This group have a similar chance of being a High Committer as at the initial branch and therefore represent a similar group of producers as the general sample.
Those with a low Convenience score at the second branch are drafted to the right at the second node, where they are then partitioned on their Autonomy score. Producers with a low Convenience score and a high Autonomy score end up in Node H with a 35 per cent chance of being a High Committer. Producers with a low Convenience score and a low Autonomy score are then partitioned based on their score for the level of Trusting relationship. Those with a high score for Trusting relationship are allocated to node I and have a 48 per cent chance of being a High Committer. Those with a lower Trusting relationship score are allocated to node J and have a 74 per cent chance of being a High Committer. It is of interest that producers with higher Trusting relationship scores are less likely to be High Committers. It may be that those that have a high commitment consider the relationship less
important because they are more concerned with other aspects of the transaction. Producers that are Low Committers may consider the relationship aspect as a more important part of the transaction.
“I’ve got a very good relationship with my buyer – when he says it’s going to happen, it happens – either to get space or move stock when needed.” Producer, Southland, South Island
While there are 10 final end nodes, the model illustrates the high influence of a producers PCA scores for Market, Convenience and Autonomy on the propensity of a producer to be a high or low committer. Not only do the first two determine the drafting from the first two branches, but they also have an effect further down the model at other branches as well. In general the higher the Market score, the more likely the producer is a High Committer. At the second branch, there is a 35 per cent change that a producer is a High Committer. If a producer has a high Convenience score they move to the Market branch and have a 29 per cent chance, while a low Convenience score goes to the Autonomy branch and has a 53 per cent chance. This illustrates that high Convenience scores reduce the chance that a producer is a High Committer, even though the opposite rule applies further down the tree. Producers with a high Autonomy score are less likely to be High Committers as illustrated at the two branch splits based on Autonomy.
High Committers are more likely to have low Convenience and Autonomy scores. These components are orthogonal, and it could be that producers with low scores for both do not value the use of a buyer to organise sales, but do not have a strong desire to have full control of decisions themselves either. This is illustrated through their higher level of commitment which places increased restrictions on their selling activities and reduced their control relative to Low Committers. Variables of positive importance in the Convenience and Autonomy components are the belief in the importance of committing to one company, and selling to a cooperative. So these cannot have a major influence on being a High Committer. Negative loadings for the two components include the perception that prices are different between companies and that there are differences between companies. These beliefs are therefore important in differentiating High Committers from Low Committers.
Given that there is a 40 per cent chance from the top branch that a producer will be a High Committer it is particularly interesting to analyse nodes where this percentage differs considerably. The first instance is at node A, where a high Market score remarkably increases the chance of being a High Committer. Similarly at node B a very low Market score significantly reduces the chance of being a High Committer. At node C with a medium Market score, medium-high Convenience, high Autonomy, and a low Influencer score, there is almost no chance of being a High Committer. At the branch for node F and G a high Convenience score increase the chance of being a High Committer. At the branch for node I and J, a lower Trusting score increases the chance of being a High Committer.